osborn@0.9.217
Voice AI coding assistant - local agent that connects to Osborn frontend
AI Security Review
scanned 27d ago · by lpm-firewall-aiThis is a user-launched voice coding assistant with no npm install lifecycle hook and no self-install chain. Local Claude credential and session paths support the user's own agent session rather than a hidden export.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
The user runs the osborn or osborn-recall command.
Impact
No unconsented credential export, remote payload, install hook, or foreign agent-config write was established from the inspected entrypoints.
Mechanism
The process checks the local Claude login, reads local session and skill files, and launches the package's own recall entrypoint.
AI rationale
This is a user-launched voice coding assistant with no npm install lifecycle hook and no self-install chain. Local Claude credential and session paths support the user's own agent session rather than a hidden export.
Decision evidence
public snapshotAI called this Clean at 86.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- package.json declares CLI bins and dev, build, and prepublishOnly scripts, with no preinstall, install, or postinstall hook and no runtime self-dependency.
- The flagged credentials constant points at the local ~/.claude/.credentials.json file used to see whether the user's own Claude login is present.
- The flagged skill reader in dist/claude-llm.js is fully commented out and is not executed.
- Project, session, and summary helpers resolve paths under the local Claude config directory, including an optional CLAUDE_CONFIG_DIR override.
- Skill listing reads SKILL.md files under the supplied local agent directory for the running assistant.
- osborn-recall is an explicit CLI that spawns this package's own local recall entrypoint.
Evidence against
- A running session can call provider APIs and, on an explicit IDE path, install code-server; those calls start only after the user launches the assistant.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
9 flagged · loading sourcebin/recall.jsView file
5L6: import { spawn } from 'child_process'
L7: import { fileURLToPath } from 'url'
High
dist/jsonl-search.jsView file
34// @vscode/ripgrep bundles the rg binary — this is our own dependency
L35: const { rgPath } = require('@vscode/ripgrep');
L36: if (rgPath && existsSync(rgPath))
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/jsonl-search.jsView on unpkg · L34dist/index.jsView file
2453const which = await new Promise((res) => {
L2454: const p = spawn('which', ['code-server']);
L2455: let out = '';
...
L2462: await new Promise((res, rej) => {
L2463: const install = spawn('sh', ['-c', 'curl -fsSL https://code-server.dev/install.sh | sh'], {
L2464: env: { ...process.env },
L2465: stdio: ['ignore', 'pipe', 'pipe'],
Critical
Same File Env Network Execution
A single source file combines environment access, network access, and code or shell execution with blocking evidence.
dist/index.jsView on unpkg · L245312setMaxListeners(50);
L13: import { createServer, request as httpRequest } from 'http';
L14: import { existsSync, readdirSync, readFileSync, mkdirSync, writeFileSync, mkdtempSync, cpSync, rmSync, statSync, utimesSync, createWriteStream } from 'node:fs';
...
L16: import { fileURLToPath } from 'node:url';
L17: import { spawn } from 'node:child_process';
L18: import { randomUUID, randomBytes } from 'node:crypto';
...
L20: import { createRequire } from 'node:module';
L21: // 0.9.71: createRequire for resolving package.json versions inside ESM
L22: const __sdkVersionRequire = createRequire(import.meta.url);
...
L24: import { PassThrough } from 'node:stream';
L25: import { createGunzip } from 'node:zlib';
L26: // Resolve __dirname for this ESM module so we can find sibling files (e.g.
Critical
Credential Exfiltration
Source appears to send environment or credential material to an external endpoint.
dist/index.jsView on unpkg · L122453const which = await new Promise((res) => {
L2454: const p = spawn('which', ['code-server']);
L2455: let out = '';
L2456: p.stdout?.on('data', (d) => { out += d.toString(); });
L2457: p.on('close', () => res(out.trim()));
...
L2462: await new Promise((res, rej) => {
L2463: const install = spawn('sh', ['-c', 'curl -fsSL https://code-server.dev/install.sh | sh'], {
L2464: env: { ...process.env },
...
L2466: });
L2467: install.stdout?.on('data', (d) => process.stdout.write(d));
L2468: install.stderr?.on('data', (d) => process.stderr.write(d));
Critical
Command Output Exfiltration
Source executes local commands and sends command output to an external endpoint.
dist/index.jsView on unpkg · L245312setMaxListeners(50);
L13: import { createServer, request as httpRequest } from 'http';
L14: import { existsSync, readdirSync, readFileSync, mkdirSync, writeFileSync, mkdtempSync, cpSync, rmSync, statSync, utimesSync, createWriteStream } from 'node:fs';
...
L16: import { fileURLToPath } from 'node:url';
L17: import { spawn } from 'node:child_process';
L18: import { randomUUID, randomBytes } from 'node:crypto';
...
L20: import { createRequire } from 'node:module';
L21: // 0.9.71: createRequire for resolving package.json versions inside ESM
L22: const __sdkVersionRequire = createRequire(import.meta.url);
...
L24: import { PassThrough } from 'node:stream';
L25: import { createGunzip } from 'node:zlib';
L26: // Resolve __dirname for this ESM module so we can find sibling files (e.g.
Critical
Download Execute
Source downloads or fetches remote code and executes it.
dist/index.jsView on unpkg · L12•matchType = previous_version_dangerous_delta
matchedPackage = osborn@0.9.205
matchedIdentity = npm:b3Nib3Ju:0.9.205
similarity = 0.781
summary = stored previous version shares package body but lacks this dangerous source file
Critical
Previous Version Dangerous Delta
This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/index.jsView on unpkgdeploy.shView file
•path = deploy.sh
kind = build_helper
sizeBytes = 2309
magicHex = [redacted]
Medium
dist/session-access.js#virtual:normalized:round1View file
•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/session-access.js#virtual:normalized:round1View on unpkgFindings
5 Critical3 High5 Medium5 Low
CriticalSame File Env Network Executiondist/index.js
CriticalCredential Exfiltrationdist/index.js
CriticalCommand Output Exfiltrationdist/index.js
CriticalDownload Executedist/index.js
CriticalPrevious Version Dangerous Deltadist/index.js
HighChild Processbin/recall.js
HighShell
HighSemantic Analysis Limiteddist/session-access.js#virtual:normalized:round1
MediumDynamic Requiredist/jsonl-search.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperdeploy.sh
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to osborn@0.9.217.
See all versions of osborn for other recorded verdicts.
- Review the evidence and your use of osborn@0.9.217 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.